Market Automator

Tech Stack Used:

Next.js 14, TypeScript, Supabase, OpenAI GPT-4, Tailwind CSS, Shadcn UI, Canva OAuth, Notion API, GitHub API, Mailchimp API

Live Link:

https://market-automator.vercel.app/

Github Repo Link:

https://github.com/Shoyo-Hin4ta/Market_Automator/

Transforming Design into Distribution

Marketing Automator bridges the gap between creative design and marketing execution. Built for small businesses and solo entrepreneurs, it's the first platform to seamlessly connect Canva designs with automated multi-channel campaign distribution.

The Challenge

Small businesses waste countless hours juggling between design tools, copywriting, email platforms, and landing page builders. The disconnected workflow leads to inconsistent branding, delayed campaigns, and lost opportunities. I identified this pain point and engineered a solution that unifies the entire process.

The Solution

I developed a sophisticated platform featuring a multi-agent AI system that acts as a virtual marketing team:

  • Orchestrator Agent: Coordinates the entire campaign creation process

  • Brand Agent: Ensures visual consistency across all touchpoints

  • Content Agent: Writes compelling, context-aware marketing copy

  • Technical Agent: Builds responsive HTML for emails and landing pages

  • Refinement Agent: Handles natural language updates like "make it more playful"

How It Works

  1. Connect & Select: Users OAuth into their Canva account and choose a design

  2. AI Analysis: The system analyzes the design's visual elements and context

  3. Content Generation: AI agents collaborate to create cohesive campaign content

  4. Preview & Refine: Users can refine outputs using natural language commands

  5. One-Click Deploy: Distribute to Mailchimp, GitHub Pages, and Notion instantly

Technical Implementation

Built with Next.js 14 and TypeScript, the platform leverages:

  • Supabase for secure authentication and encrypted API storage

  • OpenAI GPT-4 for intelligent content generation

  • Server-side rendering for optimal performance and security

  • Shadcn UI components for a polished, accessible interface

  • Real-time streaming for instant AI responses

Key Innovations

  • Visual-First AI: Unlike text-based tools, it starts with visual context from Canva

  • Multi-Agent Architecture: Specialized agents ensure professional output quality

  • Natural Language Refinement: Update campaigns conversationally

  • End-to-End Automation: From design selection to campaign analytics

Impact & Results

Marketing Automator democratizes professional marketing by reducing campaign creation from hours to minutes. It empowers businesses without technical expertise to launch campaigns that previously required entire teams and multiple subscriptions.

Technical Highlights

  • Implemented OAuth flows for 5 different platforms

  • Designed a scalable multi-agent AI architecture

  • Built comprehensive error handling and retry mechanisms

  • Created a unified API layer abstracting multiple third-party services

  • Developed real-time preview system with hot-reloading

  • Engineered secure, encrypted storage for sensitive API keys

This project showcases my ability to identify real-world problems and build innovative, production-ready solutions that combine cutting-edge AI with practical business applications.